Treatment Outcome Models in the Field

Before beginning treatment, practitioners are required to address treatment goals. Goals can be specific or broad, and range from cessation of substance abuse and regularly checking in with authorities, to compliance with medication requirements and avoiding specific locations and environments. Goals are usually developed in collaboration, with input from legal authorities, mental health professionals, and patients. Patient and community safety need to be taken into consideration as well. Goals for treatment are referred to as "outcomes." Outcomes need to be considered during treatment, during review of professional literature, and when evaluating the efficacy of treatments.

The Case of Sandy:
Sandy Lee is a 28-year-old woman who was arrested and convicted of trafficking in cocaine. As a component of her incarceration, the court required her to participate in a residential treatment program followed by outpatient substance abuse counseling when she was released from prison. During her time in the residential treatment program, she participated regularly in the group meetings and even sought individual counseling for problems associated with past abusive relationships. Shortly after Sandy completed the program, she was paroled and lived in a halfway house for approximately six months. She saw a substance abuse counselor on a weekly basis in addition to attending weekly group support sessions. Sandy also saw a psychiatrist every three months to maintain her on an antidepressant that was prescribed by the prison psychiatrist after Sandy was diagnosed with generalized anxiety disorder. Upon discharge from the halfway house, she moved into an apartment, was reunited with her children, and was able to maintain stable employment.
Four months after she had been in her own apartment with her children and maintaining a job, she was selected by her parole officer to participate in random drug testing. Three days prior to being tested, she went out on a date to a local bar. She and her date went outside and he offered her some marijuana. Although Sandy knew the risks, she also was not concerned because she had not been tested in more than three months and was certain that one "smoke" would not create problems for her. At the time, she told herself, "This is just going to let me enjoy tonight a bit more...I haven't smoked marijuana in four years and I am not planning on dealing again."
As a result of a positive drug screen three days later, her parole officer had her arrested. She was returned to jail on a "technical violation" of her parole. She went before the judge two weeks later and he revoked her parole. The judge removed her children from her custody and returned them to the care of a trusted family member. She was returned to prison and was advised by the judge to seek further treatment.

Some Questions to Ponder:
This is a real case that occurred in the not-too-distant past. Before beginning a discussion of treatment outcomes and treatment outcome models, there are some questions to consider:
1. Was Sandy's original prison stay for substance abuse treatment a success?
2. Did Sandy's outpatient treatment program result in a successful outcome?
3. Why did the judge return Sandy to prison after she tested positive for the use of marijuana? What does this say about this court's view of treatment and treatment outcomes?
4. If you were a researcher for any of the programs mentioned in this scenario, what outcomes would you focus on and measure, and how would you measure the outcomes? What would constitute success? What would constitute failure? Could there be different definitions of success and failure for different treatment models?

The Nature of Treatment Outcomes:
Treatment outcomes are important to the research question(s) being asked. Any research effort must identify how treatment success is achieved. Interestingly, this leaves the possibility that several types of outcomes are addressed in the literature using the same approach. Each person or role in a forensic setting may have a specific way in which success is determined. For instance, the court may consider success as the individual not returning, for any reason, to the justice system. The warden at the prison may consider a treatment program successful if it reduces the number of institutional offenses of the participating inmates. And the therapists in the treatment program might view success as the participant's increased frequency of contact with family members. Simply stated, each interested party in the process has his or her own view of how treatment success is defined. In addition to the "players" mentioned above (e.g., those in the courts, prisons, treatment venues, and other forensic settings), the other interested party in what determines treatment success is the person receiving the treatment services. The client might view treatment success differently from some or all of the players in forensic settings. For instance, the person who is required to participate in treatment as a condition of release from prison might view the only real desirable outcome as release from prison. Yet other individuals might have ulterior motives for treatment, such as gaining the attention of family members.
Concisely, treatment outcome is dependent on the viewpoint of the person or group being asked, "What is a desirable outcome of this treatment?" At times, there may be convergence among people in forensic settings and/or researchers on what this means, but at other times, competing research paradigms (treatment outcomes) have different definitions of success. Therefore, when the professional literature is reviewed, it is important to consider the view represented in the definition of the outcome and its success or failure. A successful outcome for one group may be different from a successful outcome of another group.
Returning to the case of Sandy Lee, treatment outcomes might be viewed as noted by the various interested parties:
1. Sandy Lee: Treatment may be viewed as a failure because she went back to prison.
2. Court: Treatment may be viewed as a success because no new charge was made for drug selling.
3. Treatment program: Treatment may be viewed as a failure due to the positive drug screen.
4. Sandy's mother: Treatment may be viewed a success because as soon as a problem was identified, she was brought back into a more structured setting for help.
5. Sandy's parole officer: Treatment may be viewed as a failure because Sandy was returned to prison.
6. Sandy's boyfriend: Treatment may be viewed as a success because their relationship improved.
The point here is that the facts of the case have not changed, only the view of what constitutes treatment success. This information is relevant not only for the researcher but also for the clinician in the forensic treatment setting. Being able to recognize the desired outcome by the particular stakeholders gives the clinician an ability to understand how competing views might define success and failure when it comes to treatment outcomes.
Three Dominant Models
In the forensic treatment professional literature, there are three dominant models of outcomes that are discussed: recidivism, relapse, and harm-reduction. These three models have a direct impact on the definition of a desired treatment outcome, how research is planned, and goals for treatment. Understanding these models not only will help clinicians understand the clinical and practice literature but also will help them plan for treatment in forensic settings.
Recidivism Model
Simply stated, and as defined in professional literature, recidivism is a person returning to prison. Although the overall notion of recidivism is the return to previous behavior patterns, the reality of the concept remains focused on the offender doing something, being caught, and then being returned to the criminal justice system. A review of numerous recidivism studies use "the return of a person to prison" as the measure of recidivism. Recidivism studies do not look at specific issues that led the person back, but look only at the situation as a binary outcome: returned or not returned.
In the case of Sandy Lee, she would be considered a treatment failure in the prison treatment program if viewed from the typical recidivism model. When incarcerated, she went through the treatment program and completed the overall program. Because she returned to the prison setting (the actual reason is not a consideration), Sandy would be considered a treatment failure according to the recidivism model.
One of the biggest drawbacks of the recidivism model is that it does not take into account why the person is returned to prison. In Sandy's case, she was returned to prison because of a technical violation of her parole (testing positive for drug use). The reason for the return is given the same weight as any reason, whether related to original reason for incarceration of not. For example, testing positive for drug abuse carries the same weight as would a murder charge.
One of the benefits of using this outcome model is that it is easy to "measure" with typical law enforcement records. Using criminal offense databases, prison records, and court documents, treatment outcomes related to recidivism may be measured without actually needing to conduct assessments of the actual individuals involved. Recidivism is easy to count and the inner rater reliability easily is established.
Relapse Model
Relapse means a return to a previous set of behaviors or mental state. The term "relapse" actually comes from the literature related to addictions and constitutes a major portion of the relapse prevention literature (e.g., Gordon and Marlatt's model) and literature related to the traditional medical model (e.g., the Alcoholics Anonymous disease model). The term "relapse" often is associated with medical and psychological models and supports the disease model. Relapse is considered part of a larger process that is unique to the individual. More importantly, the disease model and the traditional relapse prevention model relate to relapse as a normal event that needs to be addressed through treatment. It is not considered as "bad" and is seen as a part of the overall process of "recovery."
In the case of Sandy Lee, reflecting this relapse perspective, a relapse occurred when she smoked marijuana on her date. The relapse model also would suggest that there were events that led to the relapse. For instance, she may have been aware that her date had a history of using marijuana but still made the decision to go out with him. She placed herself in a relapse situation by going on the date in the first place. The fact that she was selected for the drug test and returned to prison is not relevant in the relapse perspective literature. The relapse model focuses solely on the return to previous behavior patterns or ways of behaving.
The relapse model is consistent with many of the medical and psychological models of behavior. It views behavior as cyclical and complex. The relapse prevention model has been well researched and has a strong base of data to support it. A benefit of the relapse model is that it is consistent with treatment efforts and it does not view the person as a treatment failure for just one reoccurrence of behavior, which is often referred to as a lapse. Lapses often are used in treatment as learning experiences where the client works to understand the pattern and how to prevent having a full relapse.
The difficulty of the model, however, is that it is difficult to measure accurately. In the criminal justice system and various forensic settings, there are considerable costs associated with reporting a lapse or relapse. The client who has experienced a lapse/relapse is likely to hide the occurrence out of fear of sanctions. Many forensic treatment providers are required to report relapses; therefore, the client, again, may choose to hide problems. This situation makes accurate measurement of the occurrence of any targeted behavior difficult due to the possibility of withheld information because of sanctions that would be imposed if the relapse was to be revealed.
Harm-Reduction Model
The third treatment outcome model described in forensic literature is the harm-reduction model. Of the three models, this model probably has been researched the least and is mentioned infrequently. Interestingly, it is the model that many clinicians support (in theory). According to the harm-reduction model, treatment is successful if less harm is done as a result of going through treatment as compared to no treatment at all. For example, a pedophile who goes through a treatment program for pedophilia could be considered a treatment success even if he is returned to prison for a lesser charge. Specifically, if this sex offender is returned to prison for possession of child pornography and not re-offending against an actual child, he would be considered a treatment success (less harm was done to an identifiable victim). Another example is a person who completes a program for violent behavior but returns to the treatment setting for damaging personal property without doing direct physical harm to a person. His or her aggressive behavior was reduced in terms of harmful impact to identifiable others.
When considering Sandy Lee's case, the harm-reduction model might consider her initial treatment a success because she did not return to prison for trafficking a controlled substance but only for a parole violation (not a new charge). The harm-reduction model would view success as fewer people being damaged by her return to maladaptive behaviors.
One of the primary benefits of using a harm-reduction model for treatment outcome measurement is that it may offer a better option than the "all or none" approach of the other models, and therefore may be more realistic when dealing with human behavior. It not only considers the frequency of the behavior but also takes into account the quality of the behavior. This model is reflected in some of the needle-sharing programs for heroin addicts in Europe as well as in HIV prevention programs that distribute condoms in Africa. Both of these programs are built on the harm-reduction model.
As with any outcome target, concerns are raised in the research. Some of the most popular criticisms of the harm-reduction model involve difficulty in measurement. For instance, a sex offender who does not physically harm an identifiable person by watching child pornography still is engaging in a maladaptive and deviant behavior. To say it causes less harm could be viewed as inaccurate (the children exploited in the material are victims as well). Moreover, the drug addict who uses less frequently is still at increased risk for other difficulties. Thus it is the qualitative aspect of the harm-reduction model that creates difficulty with quantitative measurement.
The Best Model
Given that these three models are evident throughout the professional literature, some have asked which is the best of the three. In fact, this is a question researchers must address when they set out to conduct a study or create an outcome measure for a given treatment plan. Moreover, since many programs are funded based on their outcomes, the choice of a model may have significant implications.
Each of these models has utility, and each offers a different perspective on a problem and its outcome. According to the author of this manuscript, a good researcher, clinician, and student considers each model in a specific situation to get an idea of the "big picture" and possible approaches to treatment and measurement of success. Said another way, each of the models allows the clinical researcher to get a view of a particular problem from a slightly different perspective, which may help in the development of an overall treatment outcome plan.
A particular benefit of understanding and being familiar with the research related to each of these models is that it supports the clinician in effectively speaking about how treatment works and helps him or her to set realistic outcomes for the client, the program, the courts, and so on. In the end, the best model is one that is used appropriately for the targeted problem or issue. This must occur in a climate that encourages an understanding, by interested parties, of the complexities of the treatment approach being utilized, and understanding how each model might or might not adequately capture the entire picture presented by the behavior in question.
Some Tips for Reading the Literature
The articles reviewed for any course in forensic treatment methods should be based on good science and research methods. As a scholar/practitioner, you should keep several key points in mind while perusing the literature. As a tool, the following questions to ask yourself are offered to assist you in gaining an appreciation for treatment outcomes.
1. What treatment outcome model is being used -- Relapse, Recidivism, Harm-Reduction, or a combination of all three?
2. Is the author presenting a limited view of the specific behavior by relying on only one model or on a model that is limited given the study? If so, what might this do to the results presented as well as the conclusions drawn?
3. Are there any risks to the author (or organization) if one type of treatment outcome model is considered? Would these risks create bias or inaccurate conclusions?
4. If the study reviewed was to be replicated using a different model, how might the results be similar and/or different?
These are four basic questions that can assist you in reading the outcome literature to provide a deeper understanding of the article as it relates to treatment outcomes.
One Final Caveat
The three treatment outcome models discussed not only have implications for the clinician and the forensic settings in which they work but also have very real meanings to the people who are undergoing the treatment. Sandy Lee presents a complicated case in that she was returned to prison for a parole violation, lost time with her children, lost the support of her boyfriend, and lost her freedom. Regardless of the model used to measure success of the treatment program, these are very real experiences for the person receiving treatment in the forensic treatment venue.
Treatment outcomes often are reported as facts in the literature. They are presented as numbers and results that remove any personal identifying information. A clinician in the forensic treatment setting, losing sight of the human cost to the client and the affected families, can become less potent as a care provider. While this does not mean a clinician acts to prevent natural and appropriate consequences for behavior, it does call on forensic treatment professionals to remain invested in understanding the individual experience of their clients and how treatment outcomes might impact their lives and the lives of those around them.

The three treatment outcome models are recidivism, relapse, and harm-reduction. Each type of outcome exists independently of the others. For example, a person may be considered a treatment success in the harm-reduction model even though he or she has had a relapse and has returned to the criminal justice system. Determining the success of treatment depends on the goals and desired treatment outcomes agreed upon at the onset of practice. Additionally, in professional literature, treatment outcomes often are indirectly defined. To comprehend the research in your field and to be an effective practitioner, it is essential to understand treatment outcome models and discern how these models are used.

describe each of the three treatment outcome models: recidivism, relapse, and harm-reduction.
- Compare (similarities and differences) the three treatment outcome models in terms of relevance in defining treatment success and/or failure with specific forensic populations, challenges in application, and advantages of each model.
- Explain at least one conclusion you drew or insight you gained as a result of your comparison